Methods for evaluating cosmetic behavior and cosmetics

Abstract

To provide a method for evaluating a cosmetic and a cosmetic behavior.SOLUTION: Disclosed is a method in which a user is requested to perform a target cosmetic behavior and a voice before use and the voice (voice after use) at the time of utterance of the user after the target cosmetic behavior is recorded. An estimated emotion value is derived from each estimated model based on the voice feature amount of the acquired voice before use and the voice feature amount of the voice after use. By comparing the derived emotion estimation values, change in the emotion of the user before and after use of the target cosmetic behavior can be analyzed.SELECTED DRAWING: None

Description

[Technical Field]

[0001] The present invention relates to a method for evaluating makeup behavior and a method for evaluating cosmetics. [Background technology]

[0002] Cosmetics have been developed with a focus on makeup habits. For example, Patent Document 1 describes a cleansing product that can reduce makeup habits. Also, for example, Patent Document 2 describes a base cosmetic product that has an excellent feel when applied to the skin, lasts well, and allows for easy makeup removal (i.e., can reduce makeup habits). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 6726344 [Patent Document 2] Japanese Patent Application Laid-Open No. 2014-221724 Summary of the Invention [Problem to be solved by the invention]

[0004] The inventors have found that changes in a user's emotions due to the use of a cosmetic product are influenced not only by the characteristics of the cosmetic product (contents) (scent, feel during use, efficacy), but also by various factors such as the appearance of the container and contents, method of use, etc. If it were possible to evaluate cosmetic behavior by comprehensively considering multiple factors without distinguishing between them, a broader perspective could be gained in the development of the cosmetic product as a whole (not only the contents, but also the appearance design and proposed method of use, etc.). Therefore, an object of the present invention is to provide a novel method for evaluating cosmetics and cosmetic behavior from the perspective of the emotions of the cosmetic user. [Means for solving the problem]

[0005] The present invention, which solves the above-mentioned problems, provides a method for evaluating a target makeup behavior, which evaluates the target makeup behavior of a user based on speech features of the user's speech before and after the target makeup behavior. According to the present invention, the user's makeup behavior can be evaluated using speech data.

[0006] In a preferred embodiment of the present invention, a change in estimated emotion value before and after the target makeup behavior of the user is calculated based on speech features at the time of speech before and after the target makeup behavior, and the target makeup behavior is evaluated using this change as an index. According to the present invention, it is possible to evaluate the influence of makeup application behavior on the user's emotions.

[0007] In a preferred embodiment of the present invention, the emotion estimation value is derived by comparing speech features of the user at the time of utterance with an estimation model that is created in advance and indicates the relationship between speech features at the time of utterance and the emotion at the time of utterance. This makes it possible to obtain emotion estimation values ​​simply and objectively using speech data.

[0008] In a preferred embodiment of the present invention, the emotion estimation value is an estimation value of any emotion selected from joy, sorrow, calm, anger, and energy.

[0009] In a preferred embodiment of the present invention, the audio features include any of the intervals between utterances, volume, pitch, talk length, speed, and frequency of pauses in speech.

[0010] In a preferred embodiment of the present invention, the cosmetic behavior includes the use of a skin care cosmetic by the user himself / herself.

[0011] The present invention also provides a method for evaluating cosmetics, which evaluates a target cosmetic behavior using the above-mentioned method for evaluating cosmetic behavior of the present invention, and evaluates the cosmetics used in the target cosmetic behavior based on the evaluation results of the target cosmetic behavior. [Effects of the Invention]

[0012] According to the present invention, it is possible to evaluate the impact of makeup behavior on a user's emotions. Furthermore, according to the present invention, it is possible to evaluate cosmetics by comprehensively understanding various elements contained in makeup behavior, which can serve as a guideline for the development of cosmetic products. [Brief explanation of the drawings]

[0013] [Figure 1] This is the first graph comparing the users' energy levels before and after using the cosmetics. [Figure 2] The second figure compares the users' energy levels before and after using the cosmetics. [Figure 3] FIG. 10 is a diagram showing the relationship between a user's satisfaction level and their level of vitality after using cosmetics. DETAILED DESCRIPTION OF THE INVENTION

[0014] In the present invention, the term "cosmetic behavior" refers to behavior that a user may take in relation to the use of cosmetics. Furthermore, "behaviors that users may take in relation to the use of cosmetics" include a series of normal actions for using cosmetics, such as opening a container, taking the contents in one's hand, taking the contents with a cosmetic tool, applying it to the skin, blending it into the skin with one's hand, removing (rinsing) the cosmetics, massaging, etc. Furthermore, the areas where cosmetics are used are not limited to the face, but can also include the entire body, including the neck, décolleté, hands, arms, and legs. For example, examples of cosmetic behavior include the use of common cosmetics, such as makeup using makeup cosmetics (e.g., foundation, eye shadow, mascara, etc.), skin care using skin care cosmetics (e.g., cleansing agent, lotion, emulsion, etc.), and sun care using sun care cosmetics (e.g., sunscreen, etc.). The cosmetic behavior to be subjected to the evaluation method according to the present invention is not particularly limited, but preferably includes the use of skin care cosmetics by the user himself / herself.

[0015] In the present invention, "emotion" refers to a feeling that arises from sensing something, and is a subjective expression that represents one's inner state and mood. Also, in the present invention, "estimated emotion" refers to an objective estimate of the user's "emotion" based on some external factors.

[0016] The method for evaluating makeup behavior of the present invention evaluates a target makeup behavior based on speech features of a user's speech before and after the target makeup behavior.

[0017] In the evaluation method according to the present invention, the target makeup behavior is preferably evaluated using as an index a change in an estimated emotion value obtained from speech features at the time of speech before and after the target makeup behavior.

[0018] In the present invention, the "speech feature" is information that can be acquired from a spoken voice. Speech features include intervals between utterances, volume, pitch, talk length, speed, and frequency of pauses in speech.

[0019] In other words, the user is prompted to speak before and after performing the target makeup behavior to be evaluated. By extracting the speech features during this utterance, it is possible to obtain estimated emotions before and after the target makeup behavior, and to observe the changes in these emotions. In this case, the emotion estimation value can be derived using an estimation model that is created in advance and indicates the relationship between speech features at the time of utterance and the emotion at the time of utterance.

[0020] Here, the "pre-created estimation model showing the relationship between speech features at the time of speech and emotions at the time of speech" can be, for example, an estimation model (multiple regression equation, discriminant equation, etc.) that is created by analyzing multiple speeches expressing different emotions and finding the correlation between speech features and emotions using statistical techniques. Furthermore, the estimation model can be an estimation model (learning model) obtained by analyzing a plurality of pieces of voice data expressing different emotions by machine learning, with the expressed emotions and the voice data as a pair. The estimation model may be a model that learns acoustic characteristics for each emotion, such as those described in Japanese Patent No. 6647722 or Japanese Patent No. 6664757. By using such an estimation model to obtain an emotion estimation value, it is possible to estimate, with a certain degree of objectivity, changes in emotions that the user does not clearly express through facial expressions or words.

[0021] In the present invention, the emotion estimated as the emotion estimation value serving as an index is not particularly limited, but preferably includes at least one of joy, sorrow, calm, anger, and energy.

[0022] Here, "energy" refers to the degree of energy that is the source of mental and physical activity required for general human activity, such as human vigor, willpower, activity, and drive. In the present invention, it is defined as a subjective relative degree of each person.

[0023] For example, if the estimated emotion values ​​of joy, energy, and calmness increase after the target makeup action, the target makeup action can be evaluated as good. Furthermore, if the estimated values ​​of sadness and anger decrease after the target makeup action, the target makeup action can be evaluated as good. Furthermore, if the direction of change in the estimated emotion value is opposite to that described above, the target makeup behavior can be evaluated as not favorable. Furthermore, from the perspective of evaluation in the development of cosmetics (products) that are intended for a wide range of users, it is preferable to use estimated values ​​of energy. This is because, while users have a variety of purposes for applying makeup (to become beautiful, to feel refreshed, to feel relaxed, etc.), an increase in energy can be positioned as a positive emotion in relation to any of these purposes.

[0024] A system (software program) for obtaining emotion estimation values ​​from voice data, including an evaluation model for emotion estimation values, can use, for example, the voice emotion analysis AI (WEB Empath API) commercialized by Empath Inc.

[0025] The evaluation method of the present invention can be carried out, for example, by the following method. First, the user's voice (pre-application voice) is recorded before applying the target makeup. The pre-application voice is not particularly limited in content or length as long as it can acquire voice features in the voice analysis process described below. For example, a voice (pre-application voice) in which the user speaks a preset phrase may be acquired, or a voice (free speech voice) in which the user freely answers (speaks) a question may be acquired. Preferably, both standard speech and spontaneous speech are acquired.

[0026] Next, the user is asked to perform the target makeup action. Next, in the same manner as the above-mentioned pre-application voice, the voice of the user speaking after the target makeup application behavior (post-application voice) is recorded. The recording method and content of the recorded voice can be the same as those described above for the pre-application voice recording method and content of the recording.

[0027] Next, the pre-use speech and post-use speech recorded above are each analyzed to obtain speech features. The method of analyzing speech features is not particularly limited and can be determined according to the speech features to be obtained. For example, speech features can be obtained by obtaining and analyzing a spectrogram of the recorded speech. Furthermore, the extraction of speech features can be performed using commonly available speech analysis software.

[0028] Next, using the aforementioned estimation model, we derive emotion estimates from the speech features of the pre-application speech and the post-application speech. By comparing the derived emotion estimates, we can analyze changes in the user's emotions before and after applying the target cosmetic behavior.

[0029] By performing the above analysis on multiple users (for example, 5 or more, preferably 10 or more), it becomes possible to generalize the evaluation of the target makeup behavior.

[0030] The above method is one example, and any method may be used as long as it is possible to "compare an estimated emotion value derived from speech features acquired from pre-use speech with an estimated emotion value derived from speech features acquired from post-use speech."

[0031] In the evaluation method according to the present invention, the steps of recording the voice, analyzing the voice, extracting voice features, calculating an estimated emotion value, and analyzing changes may all be performed using different devices, or a single device may perform multiple steps (for example, analyzing the voice, extracting voice features, calculating an estimated emotion value, and analyzing changes in the estimated emotion value), or all steps may be performed using a single device. Furthermore, of the above steps, the extraction of voice features, deriving an estimated emotion value using an estimation model, and analyzing changes in the estimated emotion value can be performed by using a computer program to cause a general-purpose computer to function. [Example]

[0032] The present invention will be described in more detail below based on examples, but the technical scope of the present invention is not limited to the following examples.

[0033] <Test example> The cosmetic behavior of using a group of cosmetics (cleansing agent, facial cleanser, lotion, and moisturizer) was evaluated based on the change in the user's estimated emotion value as a result of using the group of cosmetics consecutively.

[0034] The users were 10 women aged 35 to 49.

[0035] <Evaluation method> (1) A questionnaire survey was conducted on 10 users. All users were wearing makeup. (2) For each user, pre-use speech was recorded. Two types of pre-use speech were obtained for each user: standard speech in which the user spoke several standard short sentences, and free speech in which the user freely answered (spoke) questions based on the questionnaire survey conducted in (1) above. (3) Ten users were asked to use one of the above cosmetic products. Specifically, each user removed their makeup with a cleanser and then washed their face with a facial cleanser. Then, each user used a cosmetic product and a moisturizer. (4) For each user, post-use speech was recorded. Two types of post-use speech were obtained for each user: standard speech in which the same multiple standard short sentences as in (2) above were spoken, and free speech in which the user freely answered (spoke) about their experience using each of the group of cosmetics used in (3) above. (5) Separately, a questionnaire survey was conducted among users regarding their level of satisfaction after use of the facial cleanser, lotion, and moisturizer. The questionnaire included items about various elements of the cosmetics, and the level of satisfaction after use was calculated by scoring these items. (6) The acquired pre- and post-use speech was analyzed using speech analysis software from Empath, Inc. to obtain speech features. The speech features were the interval between utterances, volume, pitch, talk length, speed, and frequency of pauses in speech. (7) For each user, an estimated emotion value of “energy” was calculated before and after using the target cosmetic product based on the voice features obtained in (6). The analysis results are shown in Figures 1 to 3.

[0036] (Result 1) Comparison of vitality levels before and after applying makeup For the standard speech audio, no change was observed in the estimated vitality level before and after use. However, analysis of the spontaneous speech audio revealed that the estimated vitality level of the user after using the target cosmetics was higher than the estimated vitality level of the user before using the target cosmetics (i.e., the cosmetic activity of using the above group of cosmetics tends to improve the user's vitality level) (Figure 1). Furthermore, when the analysis was limited to users aged 43 to 45 (5 users), both standard speech and free speech showed that the use of the target cosmetics improved vitality levels (Figure 2). In other words, the cosmetic behavior of using the above-mentioned group of cosmetics, or the above-mentioned group of cosmetics, can be evaluated as being good because they tend to improve the vitality of the person who uses them. From the above, it has become clear that the evaluation method according to the present invention makes it possible to evaluate makeup behavior and cosmetics based on the voice obtained from the user's speech.

[0037] (Result 2) Relationship between satisfaction and estimated emotions after using cosmetics Figure 3 shows the relationship between the results of the questionnaire survey conducted in (5) of the above <Test Method> and the estimated values ​​of the users' vitality levels. For both users who were highly satisfied after using each cosmetic product ("high" in Figure 3) and users who were less satisfied after using each cosmetic product ("low" in Figure 3), the difference in vitality level before and after applying makeup (vitality level after using the cosmetic product - vitality level before using the cosmetic product) was a positive value. In other words, it was revealed that there was a tendency for vitality levels to improve as a result of using the above group of cosmetics. From the above, it has become clear that the evaluation method according to the present invention makes it possible to evaluate target cosmetics while taking into account latent factors that are not apparent in questionnaire surveys, such as user satisfaction. [Industrial Applicability]

[0038] The present invention can be applied to the development of cosmetics.

Claims

1. The method includes: a computer calculating a change in an estimated emotion value of a user before and after the target makeup application based on speech features of the user when the user speaks before and after the target makeup application; and evaluating the target makeup application using the change as an index; A method for evaluating makeup behavior, wherein the voice features are features obtained for each of a standard speech voice, which is a voice in which the user is made to speak a predetermined standard phrase, and a free speech voice, which is a voice in which the user is made to freely answer a certain question.

2. The method for evaluating makeup behavior according to claim 1 , wherein the emotion estimation value is derived using an estimation model that is created in advance and indicates a relationship between speech features at the time of utterance and emotions at the time of utterance.

3. 3. The method for evaluating makeup behavior according to claim 1, wherein the estimated emotion value is any one of estimated emotions selected from joy, sadness, calm, anger, and energy.

4. 4. The method for evaluating a makeup behavior according to claim 1, wherein the audio features include at least one of an interval between utterances, a volume, a pitch, a talk length, a speed of utterances, and a frequency of pauses in utterances.

5. The method for evaluating cosmetic behavior according to claim 1 , wherein the cosmetic behavior includes the use of a skin care cosmetic by the user himself / herself.

6. Evaluating a target makeup behavior using the method for evaluating makeup behavior according to any one of claims 1 to 5, The cosmetic evaluation method evaluates the cosmetic used in the target cosmetic behavior based on the evaluation result of the target cosmetic behavior.

Images